How come "Hardware" VA's 16 years ago sounded so good like the AN1x and now....

[ObsoleteAcc99]

[DELETED]

[Teksonik]

and a bunch of instances of some synth in mono mode

But then you'd just end up with a bunch of mono synths layered not a polyphonic synth.....

Hmmm let's see, if you loaded 61 instances of a synth or (128), limited each one's midi range to a single note, assigned each instance to it's own mixer track then added a different effect to each mixer track that might work.......

Come to think of it that would be possible in Orion but I'm not going to try.........I'll be right back......

Doh that wouldn't work......you'll just end up with per note effects not per voice effects (and I don't mean voice like Yamaha uses the term).

Ok I just had to try the experiment..I've loaded up 5 instances of DUNE in Orion and limited each one to a respond to only a single note (starting at #36). Orion automatically assigns each instance it's own mixer strip so now I could aply a separate effect for each note......and I could keep loading instances.......but then I'm thinking why? Would you ever want each note on your keyboard to have it's own effect? . And if I want to change patches I'll have to change each one individually......or I could have a different patch on each note. Cool idea or waste of time? You decide. I'm not sure either way yet........

[chroma]

so, you need one of two things: either a midi spillover VST, or a round-robin chain type thing (think hocketing, w.carlos-style).

spillover is supported by some old midi synths... let's say you wanted to play a line that might use 10 voices on a matrix 6r, which only has 6 voices. no problem - you just need two of them, you hook the second matrix 6r to the output of the first (not thru, output). when you are playing less than 6 voices, the first synth plays it. however, if you already have 6 voices playing and ask for another one, it sends it on through to the output, and the second 6r plays it.

make these all a single voice with spillover, and more than two, and you get what i was talking about. if it was a midi vst that did this, then you could put anything behind it theoretically, though the midi routing may get messy. note that if you had a multitimbral synth that allows per-patch effects processing (of which there are many) you could do a similar thing with fewer instances of the synth.

when you play a single note, it goes to the first instance and plays there. if it's held down and a second note played, it drops through to the next. not sure what the latency would be, but shouldn't be bad if this is all inside the host (via hw midi, over cables, between several 6rs, latency probably gets very high! advantage is this is a lot closer to the way voices are reused on vintage synths, so with judicious use of minor patch changes you might be able to get something more vintage sounding this way (like the way the op-x allows per-voice filter cutoff and voice tune offsets).

an alternative would be something like the ableton chain selector (which unfortunately can't be used in this way), which could just step through the chains on every note-on. this is more like hocketing, but would guarantee you'd use all the different chains eventually, however many you had. in ableton, you can attach something like this to a macro knob, and even automate it (there are people with 128 kick drums behind a knob which selects one), but no way to automatically step on midi note afaik.

i don't think i have anything offhand that can do either of these, but neither one is theoretically very complicated.

[Teksonik]

And after all that........would you really want per voice effects? Would be hard to control in real time playing.......you want delay but only on every fourth note.....only reverb on every third note played etc.....I'm not sure I'd ever use such a thing and maybe that's why nothing exists yet in vst land that could accomplish the task.......

[zerocrossing]

And after all that........would you really want per voice effects? Would be hard to control in real time playing.......you want delay but only on every fourth note.....only reverb on every third note played etc.....I'm not sure I'd ever use such a thing and maybe that's why nothing exists yet in vst land that could accomplish the task.......

Not entirely true. Isn't that what the new delay from XILS lab does? What Artillery can do?

[chroma]

plus, you'd actually have to take the time to do the programming for each voice... i can't see a reason for it either, but it is possible.

[Teksonik]

And after all that........would you really want per voice effects? Would be hard to control in real time playing.......you want delay but only on every fourth note.....only reverb on every third note played etc.....I'm not sure I'd ever use such a thing and maybe that's why nothing exists yet in vst land that could accomplish the task.......

Not entirely true. Isn't that what the new delay from XILS lab does? What Artillery can do?

I don't know....I don't use either one.......So you play a note on your synth and they route it to one effect, you hit the second note and they route it to another effect and so on? What if you hit two or three notes at the same time?

[SJ_Digriz]

I think people suffer from retroactive hearing. The majority of those VA's sounded like absolute shit. Also, you played them through external PA's and/or amps. Comparing that sound ... even the good ones ... to playing a VSTi through a DAC + monitor/headphones is about as Apples vs Truck Tires as you can get.

[Teksonik]

plus, you'd actually have to take the time to do the programming for each voice... i can't see a reason for it either, but it is possible.

I could see using it for slight pitch or filter variations on each note as you said but I'm not a big fan of chasing after pure "vintage" sounds so it would be lost on me. Actually my little experiment I touched on a few posts ago might be fun if I can get each instance of DUNE to respond to Midi Program Changes I could then have per note effects.....not the same as per voice but might be interesting just for experimental reasons.

EDIT: OK so I got each instance of DUNE to respond to Midi Program Changes.....so now I could have a separate instance of DUNE for each note on my midi controller and of course a separate effect for each note. Since only a few notes will be playing at any given time the CPU demand shouldn't be too bad......now I just need to think of a reason to have a separate effect for each note.....

[Teksonik]

I think people suffer from retroactive hearing. The majority of those VA's sounded like absolute shit. Also, you played them through external PA's and/or amps. Comparing that sound ... even the good ones ... to playing a VSTi through a DAC + monitor/headphones is about as Apples vs Truck Tires as you can get.

I agree 100%.....I pump my vsti's through 12 inch Cerwin Vegas and have active monitors running at the same time.....very unorthodox but the sound is simply amazing. I invite anyone who thinks soft synths don't have any bottom end to come over and have their teeth rattled......

[chroma]

probably no one will use it... until some guy gets some brilliant idea, creates some new sound and spawns a new subgenre. never know

i could see doing this for 'auto-pad' sounds - ex. i do sound for church, and they use the layered 'piano + pad' sound a lot. it would be nice, for example, if you could set it so that notes only played the pad if they were within 1/8 note from the beginning of a bar, for instance, or part of a chord (i.e. if more than one note is pressed in a period of <1/8 note, all those notes are considered part of the chord; notes after 1/8 bar, or some number of ms, are single again). as it is, you have to be careful or you can really make a mess with the pad underneath with the higher notes you're playing. this may already exist, i dunno, but it's not that simple because you have to hold the note until you see if another one comes in within the time frame, and then send both the new one and the saved one (along with any new ones). probably simple in max though, if i ever get it.

[Lotuzia]

so, you need one of two things: either a midi spillover VST, or a round-robin chain type thing (think hocketing, w.carlos-style).

spillover is supported by some old midi synths... let's say you wanted to play a line that might use 10 voices on a matrix 6r, which only has 6 voices. no problem - you just need two of them, you hook the second matrix 6r to the output of the first (not thru, output). when you are playing less than 6 voices, the first synth plays it. however, if you already have 6 voices playing and ask for another one, it sends it on through to the output, and the second 6r plays it.

make these all a single voice with spillover, and more than two, and you get what i was talking about. if it was a midi vst that did this, then you could put anything behind it theoretically, though the midi routing may get messy. note that if you had a multitimbral synth that allows per-patch effects processing (of which there are many) you could do a similar thing with fewer instances of the synth.

when you play a single note, it goes to the first instance and plays there. if it's held down and a second note played, it drops through to the next. not sure what the latency would be, but shouldn't be bad if this is all inside the host (via hw midi, over cables, between several 6rs, latency probably gets very high! advantage is this is a lot closer to the way voices are reused on vintage synths, so with judicious use of minor patch changes you might be able to get something more vintage sounding this way (like the way the op-x allows per-voice filter cutoff and voice tune offsets).

an alternative would be something like the ableton chain selector (which unfortunately can't be used in this way), which could just step through the chains on every note-on. this is more like hocketing, but would guarantee you'd use all the different chains eventually, however many you had. in ableton, you can attach something like this to a macro knob, and even automate it (there are people with 128 kick drums behind a knob which selects one), but no way to automatically step on midi note afaik.

i don't think i have anything offhand that can do either of these, but neither one is theoretically very complicated.

You can do this with the Synthix. Just experiment with the play modes and allocate polyphony in a wise way.

[zerocrossing]

And after all that........would you really want per voice effects? Would be hard to control in real time playing.......you want delay but only on every fourth note.....only reverb on every third note played etc.....I'm not sure I'd ever use such a thing and maybe that's why nothing exists yet in vst land that could accomplish the task.......

Not entirely true. Isn't that what the new delay from XILS lab does? What Artillery can do?

I don't know....I don't use either one.......So you play a note on your synth and they route it to one effect, you hit the second note and they route it to another effect and so on? What if you hit two or three notes at the same time?

Well, you'd probably get all the effects on all the notes. Not sure, I'm a Turnado user as I'm primarily a guitar player, but I probably could find some fun uses for Artillery too.

http://sugar-bytes.com/content/products ... hp?lang=en

and LeMasque:

http://www.xils-lab.com/pages/LeMasqueD ... rview.html

[Lotuzia]

And after all that........would you really want per voice effects? Would be hard to control in real time playing.......you want delay but only on every fourth note.....only reverb on every third note played etc.....I'm not sure I'd ever use such a thing and maybe that's why nothing exists yet in vst land that could accomplish the task.......

Not entirely true. Isn't that what the new delay from XILS lab does? What Artillery can do?

Depends if he meant every 4th notes ( one note every 4 played), or fourth beat.

If its on the 4rth beat, it can be made made in a flash in the Xils delay and also in Oxium ( see my init patch "delay when I want" in Oxium, with a delay only on the 4rth beat )

If you can predict where the 4rth note will be played in the bar, well you're again in previous case so its easy.

If, however, you cant predict it, then you have to play with Masks controling alternatively the different oscillators, together with one of the Play modes. but the game will be reduced to two, as there are only two oscillators in Oxiumn even if they can cumulate waveforms. It can produce interesting things though, are you can freely draw the masks.

With the Synthix its easier : Just use one of the polycircular modes, and program different sounds in the different layers (up to 8 ). It wont work with FX though, are fx are global, and not per layer, butit will work with everything else ( Works a bit like the Korg monopoly but more fun with more layers ) If you want total fun and unpredictable things, and a bit of suspense, just use the polyrandom mode.

[JoeCat]

They were entirely digital and ran on very, very primitive processors. You probably wouldn't even notice the amount of CPU a Virus, a AN1x, a JP-80x0 etc would take on a modern CPU.

The problem is that the analog emulation mania has gone nuts. I really had high hopes that it would end one day but it get's worse and worse instead.
The Nord Lead is still one of the best soundding VAs imo for example.

Sadly everything now needs 0df stuff and so on. That's cool and maybe even necessary if the synth has to sound as close to analog synths as possible but it's NOT (!!!!!!) necessary for a GOOD sound!

I already gave up ranting about this a long time ago though
The majority of people still want emulations and no new stuff, so I guess it will stay like that for a while.

Stuff like Zebra, Synthmaster and Diversion are the way to go imo.
The best example is Sean from ValhallaDSP. He knows a lot of stuff about the great stuff from the past but he only uses this knowledge to learn from it and to come up with unique concepts instead of completely emulating something.

Cheers
Dennis

I'm late to this party and not going to read the whole thread, but I gotta agree. I like some of the analog-modelled "stuff", but for some, it just doesn't make a difference. And some of the "authentic" sounds people (mostly outside of KVR) are falling over are either unabashedly digital (FM8, anyone?) or emulations of such. Depending on the project, digital often sounds better.

I appreciate the work a lot of devs are doing (can wait for that CS-80 emulation that's rumored to be coming by a certain well-known dev!); but if everyone's development resources are pulled in that direction there's less focus on doing something original, like Razor

(and I'm not too happy with Propellerheads for developing yet another format for developers to have to turn their attention to, but that's another thread...)